Instructions to use kandinsky-community/kandinsky-3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use kandinsky-community/kandinsky-3 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("kandinsky-community/kandinsky-3", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
| { | |
| "_class_name": "Kandinsky3UNet", | |
| "_diffusers_version": "0.24.0.dev0", | |
| "attention_head_dim": 64, | |
| "block_out_channels": [ | |
| 384, | |
| 768, | |
| 1536, | |
| 3072 | |
| ], | |
| "cross_attention_dim": 4096, | |
| "encoder_hid_dim": 4096, | |
| "groups": 32, | |
| "in_channels": 4, | |
| "layers_per_block": 3, | |
| "time_embedding_dim": 1536 | |
| } | |